1. Field of the Invention
The present invention relates to an environment recognition device and an environment recognition method for recognizing a target object based on a luminance of the target object in a detection area.
2. Description of Related Art
Conventionally, a technique has been known that detects a target object such as an obstacle including a vehicle and a traffic light located in front of a subject vehicle for performing control to avoid collision with the detected target object and to maintain a safe distance between the subject vehicle and the preceding vehicle (for example, Japanese Patent No. 3349060 (Japanese Patent Application Laid-Open (JP-A) No. 10-283461)).
Further, in such techniques, there is a technique that performs more advanced control. Specifically, it not only specifies a target object simply as a solid object, but further determines whether the detected target object is a preceding vehicle that is running at the same speed as the subject vehicle or a fixed object that does not move. In this case, when the target object is detected by capturing an image of a detection area, it is necessary to extract (cut out) the target object from the captured image before specifying what the target object is.
For example, in the case in which the captured image is a color image, there may be a method for grouping a set of pixels having a same luminance (color) and recognizing a light source such as a traffic light as a target object.
However, when an imaging device that captures a target object recognizes the target object through a windshield, and strong environment light such as sunlight (backlight) is irradiated from an imaging direction with the windshield being fogged, an entire captured image may be whitish as a whole. therefore, the color phase of the light source such as a traffic light may change in some cases, thereby deteriorating the accuracy of specifying the target object.